In general, a digital video signal is compressed and encoded according to a specific method because it has a very high capacity of data and then transmitted or is recorded on a recording medium. A Moving Picture Experts Group (MPEG) method has recently been known as the standard method of a compression and encoding system. In the MPEG method, a digital video signal is compressed and encoded using a Discrete Cosine Transform (DCT) and motion compensation. Furthermore, the compression ratio of data is increased by using variable length encoding.
A data structure in the MPEG method consists of a block layer, a macro block layer, a slice layer, a picture layer, a GOP layer, and a sequence layer in the sequence from a lower layer to an upper layer. The block layer is composed of a DCT block that is a unit for performing DCT processing. The macro block layer consists of a plurality of DCT blocks. The slice layer consists of a header part and one or more macro blocks. The picture layer consists of a header part and one or more slices. One picture corresponds to one screen.
The GOP layer consists of an Intra-coded (I) picture based on a header part and intra-frame encoding, Predictive-coded (P) pictures based on predictive encoding, and Bi-directionally predictive coded (B) pictures. The I picture can be decoded by only its own information. The P and B pictures are a reference picture. The P and B pictures require previous or subsequent pictures, and they are not solely decoded. For example, the P picture may be decoded using an I picture or a P picture, temporally earlier than itself, as a reference picture. Furthermore, the B picture is decoded by using two sheets of I pictures or P pictures before and after the B picture as a reference picture. A group, configured to include a sheet of the I picture and completed in itself, is called a Group Of Picture (GOP). The GOP becomes a minimum unit that is accessible in the streams of MPEG.
When the above compression image is streamed from a server to a terminal, a viewer who watches the image through the terminal moves to a desired scene using a fast forward or rewind function and watches the desired scene.
In order to satisfy the need of the viewer, in a conventional streaming method, the position of an I frame (or an IDR frame-H.264 reference) is recorded on the server, and transmission is performed starting with an I frame (or an IDR frame) nearest to a play position requested by the user.
The I frame (or IDR frame), however, is disadvantageous in that it has low compression efficiency if a larger number of the I frames or the IDR frames are put into a video stream at the time of encoding for the purpose of random access because it has low compression efficiency.